Angular Momentum and the Formation of Stars and Black Holes Richard B. Larson Yale University

Conclusions Compact objects such as stars and black holes generally form in larger systems such as binaries, multiple systems, star clusters, or galaxies. Gravitational interactions in these larger systems play a major role in the formation of compact objects by transporting angular momentum. The formation of objects like stars and black holes is then a much more complex, dynamic, and chaotic process than in standard models. Gravitational interactions tend to couple the mass of a forming object to the mass of the system, and this may have implications for the mass ratios in binaries, the upper IMF in clusters, and the masses of the central black holes in galaxies.

The angular momentum problem The specific angular momentum of a star-forming cloud core is 3 orders of magnitude more than the maximum that can be contained in a single star, even rotating at breakup speed. The specific angular momentum of matter in a galactic bulge is 4 – 5 orders of magnitude more than can be contained in a maximally rotating black hole. Questions: Where does the excess angular momentum go? Into residual gas, other stars, or other objects (e.g. planets)? How does it get there? What processes transport it? What is the role of magnetic, gravitational, and pressure forces?

Can magnetic fields solve the problem? Magnetic braking can remove angular momentum from diffuse clouds, but not from their dense collapsing cores where the field decouples from the gas. Rotating magnetized outflows can remove angular momentum from a newly formed object or ionized inner disk region around it. At intermediate densities, magnetic transport is ineffective and the angular momentum of the gas must still be reduced by more than 2 orders of magnitude by other effects. Machida et al 2007

The standard model Some problems: Most of the angular momentum goes into an accretion disk in which it is transported outward by an assumed intrinsic ‘disk viscosity’. Shu, Adams, & Lizano 1987 Some problems: The assumed ‘viscosity’ is problematic because all known transport mechanisms depend in some way on external circumstances (e.g. ionization). A very large disk is needed and transport times are very long (103 to 106 rotation periods) because disks are fragile and cannot sustain a large torque. Most disks probably do not last this long before being disrupted by violent interactions in a realistic system of forming stars because . . .

Most stars form in binary and multiple systems About 30% of M stars, 50% of G stars, and >70% of O stars have binary companions. This mass dependence is expected if most stars form in multiple systems that decay and preferentially eject low-mass stars. The binary frequency in some star-forming regions is up to twice that in the field. Most young stars are also found in groups and clusters, possibly in a fractal-like hierarchy. Larson 1982

3D simulations of star formation also yield many binary and multiple systems: Larson 1978 Bate, Bonnell, & Bromm 2003

The simulated binaries resemble observed ones: Bate, Bonnell, & Bromm 2002

Simulations including radiative heating show that tidal interactions are important in driving accretion onto forming stars: Bate 2009b

Simulations including radiative heating show that tidal interactions are important in driving accretion onto forming stars: Bate 2009b

Tidal interactions transfer angular momentum from circumstellar disks to stellar orbital motions, driving time-dependent accretion. Bate 2000 Krumholz et al 2009

Mass ratios in binaries Tidally-driven accretion couples the masses because the less massive star tends to accrete more rapidly and the masses tend to equalize. For close binaries, the distribution of mass ratios is roughly flat, implying a strong preference for nearly equal masses. Mazeh et al 2003

Massive stars form in denser environments with more and bigger companions: The binary frequency of massive stars approaches 100%, and their companions are typically massive and close. Orion Trapezium Massive stars are also often in Trapezium-like multiple systems and are strongly concentrated in clusters and associations. In simulations and observations, the most massive stars are often at the cluster centers: θ1C Ori Larson 1978 Muench et al 2002

“Core collapse” or “competitive accretion”? Maybe both. Simulations of the collapse of massive cores with radiative heating yield interacting binary systems, just as with low-mass stars: 8 × zoom Krumholz, Klein, & McKee 2007

Radiation pressure makes bubbles but doesn’t stop collapse. Massive interacting binaries form even when radiation pressure is included: Radiation pressure makes bubbles but doesn’t stop collapse. Krumholz et al 2009

Simulations of cluster formation reproduce the observed clustering of young stars and show massive stars forming in dense regions: They also yield a realistic-looking stellar IMF: Salpeter Bonnell, Bate, & Vine 2003

Cluster formation resembles galaxy formation Bate 2009a

Bigger clusters make bigger stars The mass of the most massive star in a young cluster increases with cluster mass roughly as Mmax ~ Mclustern , n ~ 0.5. Larson 1982 Nstars0.5 Mcluster0.5 Mcluster0.74 Maschberger & Clarke 2008 Weidner & Kroupa 2006

This can produce a power-law upper IMF: If stars form in a clustering hierarchy and the mass of the most massive star in each group increases as a power n of the group’s mass, a power-law IMF results: dN/d log m ~ m−x, x = 1/n A Salpeter IMF (x = 1.35) results if n = 0.74, possibly consistent with observations for masses below 30 Mʘ (Weidner & Kroupa 2006). The mass of the most massive star in a cluster should increase with cluster mass because a bigger cluster can redistribute more mass and angular momentum. In both binary systems and clusters, the mass of the most massive object may then be coupled to the mass of the system. This seems to be true also in galaxies . . .

Bigger galaxies make bigger black holes The masses of central black holes and stellar nuclei in galaxies scale roughly with bulge mass as MBH ~ 0.0015 Mbulge black holes stellar nuclei Ferrarese et al 2006

Black hole building can occur by merging or by gas accretion: Merging of the central black holes of merging galaxies can occur by a combination of large-scale dynamical friction and small-scale gravitational drag effects: 120 kpc 8 kpc ~ 100 pc ~ 1 pc 60 kpc 160 pc Mayer et al 2007 Escala, Larson, Coppi, Mardones 2005

Black hole growth by gas accretion Standard accretion disks become inefficient and unstable at radii > 0.1 pc, so if gas is to get into such a region from a galactic bulge, its angular momentum must be reduced by at least 3 orders of magnitude in some other way. Some possibilities: A more extended gravitationally unstable gas disk may develop spiral features and gravitational torques that drive inflows. Massive clumps may also form and lose angular momentum by gravitational drag and fall inward. Galactic bars and disk asymmetries can exert torques on the gas and drive inflows on a large range of scales. All of these phenomena could be consequences of galaxy interactions and mergers, and all are seen in simulations.

Evolution of a massive nuclear gas disk A massive unstable gas disk like those observed in ULIRGs can fragment into clumps and filaments that simultaneously form stars and drive an inflow. Escala & Larson 2008 The dynamics of such a gas disk can be violent and chaotic, and can lead to comparable rates of star formation and black hole feeding. 500 pc Escala 2006 (unpublished) Escala 2007

Massive clusters and star formation near the Galactic Center Hubble-Spitzer Galactic Center mosaic 10 pc Arches cluster Quintuplet cluster Galactic Center Pistol star HST NICMOS 1.87 μm (Paα) + Spitzer 3.6, 4.5, 5.8, 8.0 μm

Hubble-Spitzer Galactic Center mosaic molecular clouds 2 pc Galactic Center Trapezium

Compact nuclear clusters in the Milky Way and M31 Compact clusters of massive young stars less than 1 pc in size surround the central black holes of both the Milky Way and M31: M31 nucleus MW Sgr A* 2 pc blue cluster IRS 13 complex eccentric disk 0.7 pc Genzel et al 2003 Bender et al 2005

The Galactic Center Cluster extends to within 50 AU of the central black hole! Gillessen et al 2010 0.04 pc So whatever made these stars brought matter very close to the black hole. These stars may then be leftovers from a black-hole feeding event.

Star formation near a central black hole How did the observed massive stars form within a few tenths of a parsec of the central black holes in both the Milky Way and M31? In an accretion disk, possibly eccentric? In gas captured into orbit around the BH? Nayakshin, Cuadra, & Springel 2007 Bonnell & Rice 2008 In either case, a significant fraction of the gas may be accreted by the black hole.

Bars and asymmetric disks can also drive inflows: Bar torques can drive gas into the inner kiloparsec of a galaxy, and nuclear bars or spirals may drive inflows on smaller scales. Asymmetric nuclear disks like that seen in M31 can similarly produce torques that drive inflows in the inner few parsecs. Effect of a galactic bar Gas orbits in the inner parsec of M31 ● P2 shock Maciejewski et al 2002 Chang et al 2007

Simulations show that gas falling into a galactic nucleus forms an asymmetric disk like that in M31, and that the gravitational torque due to the disk drives continuing gas infall toward the black hole: gas stars 50 pc Hopkins & Quataert 2010 The angular momentum of the gas accreted by the black hole then ends up mostly in stars, and it is transmitted to them by gravity.

Hubble-Spitzer Galactic Center mosaic molecular clouds 2 pc Galactic Center Trapezium

log Mmax • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 10 • • • • • • • • • • • • • • • • • • 8 stellar nuclei • • • • black holes • • • • • • (Ferrarese et al 2006) • • • • • • • • • • • • • • • • • Mmax = 0.0015 Msystem • 6 • • • • low-mass black holes (Greene et al 2008) • 4 G1 (M31)? Pistol star (Gebhardt et al 2005) • θ1C Ori M15?? * close binaries * * 2 * * * * * * * * * * * * * * stars * * 2 4 6 8 10 12 log Msystem

Conclusions Compact objects such as stars and black holes generally form in larger systems such as binaries, multiple systems, star clusters, or galaxies. Gravitational interactions in these larger systems play a major role in the formation of compact objects by transporting angular momentum. The formation of objects like stars and black holes is then a much more complex, dynamic, and chaotic process than in standard models. Gravitational interactions tend to couple the mass of a forming object to the mass of the system, and this may have implications for the mass ratios in binaries, the upper IMF in clusters, and the masses of central black holes in galaxies.